Taxonomy and Biogeography of Bunopus Spatalurus (Reptilia; Gekkonidae) from the Arabian Peninsula
Total Page:16
File Type:pdf, Size:1020Kb
Accepted on 14 June 2015 © 2015 Blackwell Verlag GmbH J Zool Syst Evol Res doi: 10.1111/jzs.12107 1Institute of Evolutionary Biology (CSIC-Universitat Pompeu Fabra), Barcelona, Spain; 2Faculty of Life Sciences and Engineering, Departament Produccio Animal (Fauna Silvestre), Universitat de Lleida, Lleida, Spain; 3CIBIO, Centro de Investigacßao~ em Biodiversidade e Recursos Geneticos, Universidade do Porto, InBio Laboratorio Associado, Vairao,~ Portugal; 4Departamento de Biologia, Faculdade de Ciencias da Universidade do Porto, Porto, Portugal; 5Department of Biology, Villanova University, Villanova PA, USA; 6Department of Ecology, Faculty of Science, Charles University in Prague, Prague, Czech Republic; 7Drıtec 65, Drıtec, Czech Republic; 8Department of Zoology, National Museum, Prague, Czech Republic Taxonomy and biogeography of Bunopus spatalurus (Reptilia; Gekkonidae) from the Arabian Peninsula 1,2 1,3,4 5,† 6 6 PHILIP DE POUS ,LUIS MACHADO ,MARGARITA METALLINOU ,JAN CERVENKA ,LUKA S KRATOCHVIL ,NEFELI 1 7 8 1 2 1 PASCHOU ,TOMA S MAZUCH ,JIRI SMID ,MARC SIMO -RIUDALBAS ,DELFI SANUY and SALVADOR CARRANZA Abstract In the last decade, taxonomic studies have drastically increased the number of species known to inhabit the Arabian deserts. While ongoing phyloge- netic studies continue to identify new species and high levels of intraspecific genetic diversity, few studies have yet explored the biogeographic pat- terns in this arid region using an integrative approach. In the present work, we apply different phylogenetic methods to infer relationships within the Palearctic naked-toed geckos. We specifically address for the first time the taxonomy and biogeography of Bunopus spatalurus Anderson, 1901, from Arabia using multilocus concatenated and species tree phylogenies, haplotype networks and morphology. We also use species distribution modelling and phylogeographic interpolation to explore the phylogeographic structure of Bunopus spatalurus hajarensis in the Hajar Mountains and the roles of climatic stability and possible biogeographic barriers on lineage occurrence and contact zones in this arid mountain endemism hot spot. According to the inferred topology recovered using concatenated and species tree methods, the genus ‘Bunopus’ is polyphyletic. Bunopus tuberculatus and B. blan- fordii form a highly supported clade closely related to Crossobamon orientalis, while the two subspecies of ‘Bunopus’ spatalurus branch together as an independent highly supported clade that diverged during the Miocene according to our estimations. Within B. s. hajarensis, three geographically structured clades can be recognized that according to our estimations diverged during the Late Miocene to Pliocene. The paleodistribution models indi- cate climatic stability during the Late Pleistocene and the lineage occurrence, and predicted contact zones obtained from phylogeographic interpolation therefore probably result from the older splits of the groups when these lineages originated in allopatry. As demonstrated by the results of the multilo- cus molecular phylogenetic analyses and the topological test carried out in this study, the genus ‘Bunopus’ is not monophyletic. To resolve this, we resurrect the genus Trachydactylus Haas and Battersby, 1959; for the species formerly referred to as Bunopus spatalurus. Considering the morphologi- cal differences, the high level of genetic differentiation in the 12S mitochondrial gene and the results of the phylogenetic and the cmos haplotype net- work analysis, we elevate Trachydactylus spatalurus hajarensis to the species level Trachydactylus hajarensis (Arnold, 1980). Key words: Paleodistribution modelling – phylogeography – multilocus phylogeny – spatial interpolation – contact zone – Palearctic naked-toed gecko Introduction identify new species, especially in the southern Arabian Penin- sula (Babocsay 2004; Busais and Joger 2011; Carranza and Deserts encompass a large portion (12%) of the Earth’s land sur- Arnold 2012; Metallinou and Carranza 2013; Sm ıd et al. 2013b, face and are important for understanding global biodiversity pat- 2015; Vasconcelos and Carranza 2014) as predicted by Ficetola terns. While deserts are thought to have relatively low species et al. (2013). Likewise, several studies have revealed high levels richness compared to other biomes, they are often inhabited by of intraspecific genetic structure of Arabian lizards, mostly in many specialized species and clades possessing a wide array of species inhabiting mountainous areas such as the Hajar Moun- unique adaptations to arid conditions. For example, lizard com- tains in northern Oman and the UAE and the mountains in munities in deserts might be richer compared to warm temperate southern Oman and Yemen (e.g. Carranza and Arnold 2012; and tropical regions, and they can contain as many as 70 differ- Sm ıd et al. 2013a), but also in species occurring in lowlands ent species co-occurring at single localities (Pianka 1973; (Metallinou et al. 2012, 2015). Two biodiversity-rich areas with Rabosky et al. 2011). In the past, extensive work on desert rep- high levels of endemicity are recognized within Oman: the Hajar tiles has mainly focused on Australian, South African and North Mountains in north Oman and the Dhofar Mountains in south American deserts (e.g. Pianka 1986), while fewer studies exist Oman and east Yemen. These two mountainous regions have for the Arabian deserts (e.g. Anderson 1896; Haas 1957; Arnold their own unique and complex geologic histories (reviewed in 1972, 1980, 1986; Gasperetti 1988; Sch€atti and Gasperetti 1994; Carranza and Arnold 2012 and Gardner 2013; see also Arnold Sch€atti and Desvoignes 1999; Carranza and Arnold 2012; Gard- 1977, 1980 and other articles in the same volumes) but share dis- ner 2013). Recently, however, an increasing number of taxo- tinct climatic conditions and vegetation that differentiate them nomic, biogeographic and phylogenetic studies have focused on from the much more arid surrounding lowland desert. Although lizards inhabiting these deserts (e.g. Carranza and Arnold 2012; the Hajar and the Dhofar Mountains are inhabited by partially Metallinou et al. 2012, 2015; Metallinou and Carranza 2013; distinct reptile communities with unique endemic species (e.g. Sm ıd et al. 2013a). In the last decade, taxonomic studies have Hemidactylus, Asaccus geckos) and even genera (Omanosaura, drastically increased the number of species inhabiting the Lacertidae) (see Arnold 1986; Arnold and Gardner 1994; Car- Arabian desert, while ongoing phylogenetic studies continue to ranza and Arnold 2012), several species occur in both regions while being absent from the lowland areas between them (Buno- Corresponding author: Salvador Carranza ([email protected]) pus spatalurus Anderson, 1901, Chalcides ocellatus (Forskal, †Author deceased July 2015 J Zool Syst Evol Res (2016) 54(1), 67--81 68 DE POUS,MACHADO,METALLINOU, CERVENKA,KRATOCHVIL,PASCHOU,MAZUCH, SMID,SIMO-RIUDALBAS,SANUY and CARRANZA 1775), Trachylepis tessellata (Anderson, 1895) and Platyceps the genus Bunopus (Bauer et al. 2013), as well as three of the four spe- rhodorachis (Jan, 1865); see Gardner 2013). cies of this genus (Bunopus crassicauda Nikolsky, 1907 missing) and Among the species exhibiting this notable north–south distri- two out-groups (Hemidactylus brasilianus (Amaral, 1935) and bution pattern, Bunopus spatalurus is an interesting representa- H. haitianus Meerwarth, 1901). Given the focus of the study, sampling tive, as previously mentioned by Arnold (1980). This species is was especially intensive in Bunopus spatalurus. A list of all the speci- mens with their GenBank accession numbers, voucher codes and locality widely distributed in the southern Arabian Peninsula, ranging information is presented in Table 1. from western Yemen to the mountains of northern Oman and the Genomic DNA was extracted from ethanol-preserved tissue samples UAE, but with some areas of absence (Sindaco and Jeremcenko using the standard high salt method (Sambrook et al. 1989). Up to six 2008; Gardner 2013). Arnold (1980) described the subspecies genetic markers were PCR-amplified and sequenced in both directions Bunopus spatalurus hajarensis from the Hajar Mountain range (see Table S1 for primer details): one fragment of the mitochondrial gene and Masirah Island based on clear morphological differences. encoding the ribosomal 12S rRNA (12S; primers 12SaGekko and However, despite some recent phylogenetic studies on Palearctic 12SbGekko – Metallinou et al. 2015), and five fragments of the nuclear naked-toed geckos (Cervenka et al. 2008; Bauer et al. 2013), no genes encoding the oocyte maturation factor Mos (cmos; primers FU-F – molecular studies have yet addressed the status of Bunopus and FU-R Gamble et al. 2008), the recombination-activating gene 1 – – spatalurus. This is possibly the result of the difficulty in finding (rag1; primers F700, R700 Bauer et al. 2007; and R13 and R18 Groth and Barrowclough 1999), the recombination-activating gene 2 specimens of B. s. spatalurus compared with the relatively abun- (rag2; primers PY1-F and PY1-R – Gamble et al. 2008), the acetyl- dant B. s. hajarensis (Haas and Battersby 1959; Arnold 1980; cholinergic receptor M4 (acm4; primers int-F and int-R – Gamble et al. Gardner 2013) as well as the current political instability in 2008) and a short fragment of phosducine (pdc; primers PHOF2 and Yemen that obstructs fieldwork. Clarifying the phylogenetic posi- PHOR1 – Bauer et al. 2007). PCR conditions used for the amplification